Information Transparency,Corporate Social Responsibility,and Credit Rating: Evidence from Taiwan’s Enterprises

Abstract

Based on agency theory and overinvestment view, the research explores the possibility of severe agency problems caused by managers’ self-interest and promote information transparency to solve agency problems. Furthermore, according to the risk mitigation viewpoints, Corporate Social Responsibility (CSR) performance influences credit risk rating. Therefore, there is an interlocking relationship between the three parts, which are examined by panel data analysis. First, the Tobit regression is used to predict the enterprises’ ranking that did not win the CSR Award of Common Wealth Magazine. Second, the relationships among information transparency, CSR, and credit risk index (TCRI^TM) provided by Taiwan Economic Journal are analysed. The best model is chosen by using T test, LM test, and Hausman tests. Next, this research uses the fixed-effects model as empirical analysis. The major empirical findings of this paper are as follows: (a) Improving information transparency can promote CSR performance. (b) Improving information transparency helps obtain better credit ratings. (c) Better CSR performance helps obtain better credit ratings. In summary, all three factors positively affect a company, improving its investment environment and bringing about a good reputation. When business managers or decision-makers are developing business strategies, this could be something to take into account.

Keywords

Corporate Social Responsibility (CSR)Credit Rating Information Transparency Panel Data Regression Tobit Regression

Introduction

Transforming business models from sole proprietorships or partnerships into corporations may lead to agency problems where managers and shareholders have different interests. Previous literature has pointed out the impact of information asymmetry in the generation of agency problems (Fama & Jensen, 1983; Hart & Moore, 1995). Enterprises can reduce the negative effect of agency problems through information disclosure and announcing part of their business strategies and financial information to external investors.

Previous studies have mainly explained the empirical results from a single theoretical perspective (Cherian et al., 2020; Fabozzi et al., 2022; Park & Lee, 2021). This study combines a variety of theoretical perspectives to examine how decided by enterprise internal actions, such as information transparency and Corporate Social Responsibility (CSR) involvement, affect the outcomes of external evaluations, such as credit agency evaluations of corporate credit risk. This study also explores the manager’s complex mental journey of corporation information disclosure. This study is divided into three parts: First, information transparency is explored in terms of its impact on CSR performance. Second, this study analyses the degree of CSR involvement of internal corporate affects the credit rating of external credit rating agencies. Third, this study evaluates the complexity of the process in which a corporate manager is willing to disclose information deeply affects the motivation of enterprises to disclose information. It affects the evaluation of the enterprise’s credit by external credit rating agencies. The relationship between internal and external enterprises and the willingness of the manager to disclose information has rarely been discussed in previous studies.

The research design of this study used a combination of time-series data and cross-session data for the panel data analysis. This method has the advantage of the continuous dynamic character of time series and the advantage that cross-sectional data can express the characteristics of different samples (Hsiao, 1986). The main finding of this study is that the internal behaviour of a company affects the evaluation of the company by external organizations. The results can then persuade the managers to make decisions favourable to their corporate image.

The background of this study revolves around the Taiwan region, which is a small open economy. Therefore, a large share of the business environment comes from the export market, which is linked to globalization (Lau, 2020; Lee et al., 2010). Companies need to make an effort to implement the content of CSR and obtain relevant certifications, as well as get better credit ratings to absorb the attention of international brands (Bannier et al., 2012; Chen et al., 2013). According to the Global Reporting Initiative (2020) database, Taiwan ranked eighth in the total number of individual CSR reports from 135 economies through 2020, indicating that Taiwan is one of the examples of CSR promotion among small open economies. Taiwan has similar economic and trade conditions to most economies and developed earlier than most small economies (Hauge, 2019). Therefore, Taiwan is a worthy model as a research context for most small open economies.

Not only are the companies in Taiwan making efforts to respond to international trends, but government agencies in Taiwan are also implementing CSR-related policies. In 2010, The Financial Supervisory Commission and Taipei Exchange issued ‘Ethical Corporate Management Best Practice Principles for TWSE/GTSM Listed Companies’, in which companies are advised to follow the principle of enhancing information disclosure related to CSR. Most corporations believe that executing CSR can make them socially acceptable to the greater public (Kotler & Lee, 2005), and thus corporations consider CSR expenditures as investments finding a successful operating direction (Porter & Kramer, 2006). Fulfilling CSR is critical for differentiating a corporation from its competitors, ultimately enhancing its reputation (Aguilera-Caracuel & GuerreroVillegas, 2018; Al-Shammari et al., 2021; Lin et al., 2019; Phang & Hoang, 2021). A company’s reputation is also a major factor in the ratings from credit institutions (Pittman, 2008). A credit rating is based on operating ability, asset-liability management, and other metrics to quantize the enterprise’s credit attributes. The board of directors can maintain or improve the corporation’s credit rating if they fully understand the effect of CSR on credit rating.

Fitch Ratings (2004) pointed out that credit ratings integrate corporate governance methods to improve the evaluation process. The agency theory and the overinvestment assumption framework define corporate governance by estimating the board of directors’ ownership structure and independence from the creditors’ perspective. Corporations with bad corporate governance or information disclosure often face a downgrade in ratings or negative evaluations. Business management also has an important role in CSR implementation (Jamali et al., 2008). Based on agency theory, we explore the possibility of severe agency problems caused by managers’ self-interest based on the overinvestment view and promote information transparency to solve agency problems.

Furthermore, the impact of information transparency extends to CSR performance and credit risk rating. According to the risk mitigation viewpoints, CSR performance influences credit risk rating. Therefore, there may be an interlocking relationship between the three parts. This research looks to discover the associations between information transparency, CSR performance, and credit ratings of corporations in Taiwan to prove the following statements: (a) corporations with higher information transparency exhibit better CSR performance, (b) corporations with higher information transparency have higher credit ratings, and (c) corporations with higher CSR ratings have higher credit ratings.

Literature Review and Hypotheses’ Construction

The agency problem is caused by information asymmetry between managers and shareholders (Jensen & Meckling, 1976). Barnea and Rubin (2010) used the overinvestment view to further show that overinvestment in CSR is not for the public good to maximize shareholders’ profit but to enhance the managers’ private interests, resulting in inefficient use of corporate resources, and even make managers ignore the core issues that need attention in their original duties (Bénabou & Tirole, 2010; Djankov et al., 2008; Ferrell et al., 2016; Hassen & Militaire, 2020). In order to reduce the various costs and the public’s distrust towards the managers derived from this agency problem, managers must improve information transparency. This act can also reduce the information asymmetry between managers and shareholders (Arab et al., 2020; Diamond & Verrecchia, 1991; Keim, 1978). In order to solve these problems, managers need to make information transparent in order to reduce the outside community’s concerns. Therefore, this study infers that the higher the agency cost, the stronger the incentive for operators to make information transparent voluntarily.

Relationship Between Information Transparency and CSR Performance

There are many ways to make information transparent (Huang et al., 2022; Martínez‐Ferrero et al., 2015). The production of non-financial reports is considered the first step towards transparency by most companies. They generally hope to have in-depth communication with stakeholders to lessen the occurrence of information asymmetry (Anwar & Malik, 2020; Benlemlih et al., 2021; Brown et al., 2011). A company’s value is strengthened through information disclosure, and CSR and corporate integrity efforts can be enhanced, leading to a company maintaining sustainable developments.

Based on the aforementioned reasons, this research believes under the agency theory and the overinvestment view framework that more information a company discloses will help increase the transparency of that information, further causing the company’s agency cost to decrease. In order to gain their stakeholders’ trust, managers are typically willing to eliminate information asymmetry and fulfil their social responsibilities as corporate agents, resulting in better social responsibility by the firm. The first hypothesis can be constructed based on the above.

H ₁: The more information a company discloses, the better corporate social responsibility performance is.

Relationship Between Information Transparency and Credit Rating

Managers employ appropriate channels to disclose company information in order to increase the company’s information transparency (Choy et al., 2006; Healy & Palepu, 2001). Ashbaugh-Skaife et al. (2006) supported that transparent financial reports decrease the possibility of information asymmetry between the company and shareholders. When a company discloses more information, shareholders expect the possibility of the company hiding negative information is less, default risk decreases (Chau & Gray, 2002; Pawlina & Renneboog, 2005). Duffie and Lando (2001) suggested that incomplete disclosure of corporate information affects the credit risk of an enterprise. Companies with high information transparency help to reduce the incidence of material misjudgement by stakeholders (Ang & Brau, 2002). Ashbaugh-Skaife et al. (2006) pointed out that information transparency is a critical part of corporate governance, which can keep the company’s financial performance strong (Aldamen et al., 2020), thus reducing the risk of shareholders depriving the rights of creditors. Thus, credit rating institutions can give a better credit rating.

The managers will choose to make corporate information public in order to gain stakeholders’ trust or reduce agency costs (Sethuraman, 2019; Tahir et al., 2022; Yu et al., 2018). Corporate information can be divided into good news that is good for the future of the company and bad news that is expected to be bad for the future. Compared to positive information, companies easily delay or do not disclose negative information (Kothari et al., 2009). If managers make the good news public, it helps investors to have more confidence in their companies (Gurun & Butler, 2012; Healy et al., 1999; Milgrom, 1981). Suppose managers reveal bad news about a company’s performance that some studies suggest that this may only partially reduce trust in the company. The results of these studies show that by accelerating the disclosure of bad news by managers, stakeholders are more likely to believe in the bad news disclosure explanation and gain more trust (Hutton et al., 2003; Mercer, 2005). If managers see the long-term implications of hiding negative information, they may find that this behaviour creates serious information asymmetries that can lead to enormous litigation costs or potential claims for other violations to compensate for inadequate information disclosure or credit risk (Baginski et al., 2002; Bao et al., 2019; Chi et al., 2009). Conversely, when managers are willing to disclose negative private information, it may reduce the firm’s capital costs and losses due to fear of risk (Healy & Palepu, 2001; Skinner, 1994).

Credit rating agencies consider the quality of information disclosed by the whole company to assess the default risk of the company in order to determine the credit rating level of the company (Sengupta, 1998; Yu, 2005). Credit rating agencies can obtain confidential information about the debt issuer and use public information about the firm to assess credit risk (Bonsall, 2014; Bushman et al., 2010; Jorion et al., 2005). The rating agency assesses that if the company has a higher chance of hiding unfavourable information, the higher the chance that the company will be assessed as being at risk of default. It indicates that the company does not have a solid corporate image to stakeholders, hence lowering the credit risk rating (Ahn et al., 2019; Micu et al., 2004). Therefore, after evaluating the severity of the long-term impact of disclosing negative information, managers believe that disclosing relatively unfavourable corporate information is much less costly than concealing unfavourable information (Hong & Kacperczyk, 2010; Kumar et al., 2012). Consequently, the willingness of corporate managers to disclose more information reduces information asymmetries. Credit rating agencies can assess the reduced probability of default risk, which improves the company’s credit rating.

H ₂: A company with higher information transparency has a better credit rating.

Relationship Between CSR Performance and Credit Rating

Previous studies confirmed the value created by CSR from various theoretical perspectives, such as the risk mitigation view. The risk mitigation theory shows that enterprises engaging in CSR activities can effectively reduce business risk, leading to higher credit ratings because they provide information on the company’s default rate (Bae et al., 2018; Goss & Roberts, 2011; Jo & Na, 2012; Starks, 2009). Ye and Zhang (2011) illustrated that positive moral capital could effectively reduce stakeholders’ negative evaluations of corporate misbehaviour. Therefore, if companies are more involved in CSR activities associated with lower risk, then the impact of CSR on credit ratings should be positive (Attig et al., 2013; Jiraporn et al., 2014).

Boutin-Dufresne and Savaria (2004) found that the less a company invests in social responsibilities, the more significant its unsystematic risk will be, implying that its credit risk is severe. Godfrey (2005) showed that if the company is not committed to social responsibility activities, it may face legal and economic sanctions from stakeholders, blocking the company from external resources and increasing its business risk. The consequences are harmful to the company’s credit ratings. With the arguments mentioned above, this research infers that if a company actively invests in social responsibility activities, then its CSR grading will be higher, leading to a better credit rating.

H ₃: A company with a better corporate social responsibility performance has a better credit rating.

Research Design

Research Sample and Data Sources

Common Wealth Magazine evaluates listed companies in Taiwan with three consecutive years of record revenue and selects those with good CSR performances by considering four aspects: social participation, organizational commitment, environmental sustainability, and corporate governance. The main targets of this research are companies that won the CSR award between 2007 and 2014. The reason behind the time selection is that 2014 was the final year that the information transparency metric was listed. According to the Securities and Futures Institute, it was later merged into corporate governance evaluation. The detailed grades, however, are only partially available to the public. Only the detailed data of companies with the top rankings in the CSR award can be observed. Thus, with such a characteristic, the data are censored.

From the explanation of the Taiwan Corporate Credit Risk database provided by the Taiwan Economic Journal (TEJ), some industries with special characteristics are excluded from the credit ratings. The financial industry, for example, has unique operating standards and rules that may affect the control variables and regulations toward CSR disclosures (Fuente et al., 2017). Based on this setting, this research filters out four types of companies, including: (a) financial industry, companies such as Fubon Financial Holding Co., Ltd. and Cathay Financial Holdings Co., Ltd.; (b) life insurance industry, companies such as China Life Insurance Co., Ltd. and Cathay Life Insurance Co., Ltd.; (c) public enterprises including CPC Corporation and Taiwan Power Company; and (d) companies with incomplete financial statements. With the original data size of 712 observations, the final dataset includes 528 observations from 67 companies after the filtering process.

Financial data such as the debt ratio and return on assets (ROA) are from the consolidated finance database of TEJ. Corporate governance data, including information disclosure rankings (information transparency), are from the corporate governance database of TEJ. Company credit rating data are from the Taiwan Corporate Credit Risk Index of TEJ.

Explanatory Variables

According to the TCRI^TM Rating system, the credit ratings are divided into nine levels: levels 1–4 are the low-risk tier, levels 5 and 6 are the medium-risk tier, and levels 7–9 are the high-risk tier. This research measures CSR performance based on the CSR award of Common Wealth Magazine, which evaluates and grades companies’ CSR performances and ranks the companies from the top score to the bottom one.

Control Variable

The Size of the Board of Directors

Hillman and Dalziel (2003) believed that board members with experience could evaluate their company’s business strategies and the effect on CSR by using the right to supervise. Past research found a positive correlation between the scale of the board of directors and firm performance through empirical studies (Aggarwal & Nanda, 2004; Nwude & Nwude, 2021). This research follows the method of Barnea and Rubin (2010) to use the total number of board members to evaluate the size of the board and expects a positive correlation between board size and a firm’s CSR performance.

The Number of Independent Directors

Hermalin and Weisbach (2003) noted that outside independent directors, who have a more independent status, can supervise other directors and managers and are less likely to present any self-interested behaviour (Jensen & Meckling, 1976). Independent directors can provide effective suggestions, impartial supervision, and objective decisions, helping the company increase overall performance. This research follows the model settings of Jo and Harjoto (2011) to calculate the proportion of independent directors as the number of independent directors divided by the number of total directors and expects to see a positive correlation between this proportion and the company’s CSR performance.

Institutional Investors’ Holding Ratio

McConnell and Servaes (1995) used Tobin’s Q to evaluate company performance. They found that institutional investors’ shareholding ratio has a positive effect on company performance since a higher shareholding ratio will motivate institutional investors to supervise the decisions of company managers (Shleifer & Vishny, 1986) and give them more power to block actions that may harm the company (Brickley et al., 1988). This research follows the method of Barnea and Rubin (2010) to calculate institutional investors’ shareholding ratio variable by adding friendly legal persons’ shareholding ratio and external corporations’ shareholding ratio and expects to see a positive correlation between this variable and the company’s CSR performance.

Insiders’ Shareholding

Insiders’ shareholding represents the shareholding ratio of supervisors and directors, and professional managers of the company. The direction of insiders’ shareholding’s effect on a company’s CSR performance is unclear. The board of directors has the responsibility to supervise managers’ performance. When company performance is below expectations, the board of directors can decide about any change in management personnel. Therefore, when the board of directors owns more shares, it has a stronger connection with the company’s financial performance (Yermack, 1996) and is motivated to focus more on the supervision of company management (Berle & Means, 1932), leading to the company being more active in CSR. According to the convergence of interest hypothesis (Jensen & Meckling, 1976), the shareholding of professional managers will affect their self-interest. The more shares professional managers own, the more similar their self-interest is in line with stakeholders, leading to an increase in firm value.

Jensen and Ruback (1983) proposed an opposite idea. The entrenchment hypothesis states that when managers have a higher shareholding ratio, they have more security in their positions and may become indifferent toward stakeholders’ interests, leading to a decrease in firm value. Thus, the effect of the shareholding ratio of managers on the company’s CSR performance is ambiguous.

Company Size

Johnson and Greening (1999) presented the connection between a company’s size and CSR performance. When the company’s size is larger, the more likely will its operational activities affect the environment. For example, a large company has to internally build a safe working environment and externally be responsible to stakeholders. Fombrun and Shanley (1990) said that the public pays more attention to more prominent companies. Thus, company size affects CSR practices at some level. The larger the company’s size is, the more likely the company practices CSR activities.

Return on Assets (ROA)

Waddock and Graves (1997) and Preston and O’Bannon (1997) both used the ROA metric to measure the financial performance of a company practicing CSR. The results of these works showed a positive correlation between ROA and the financial performance of a company practicing CSR.

Debt Ratio

The debt ratio is commonly used to measure a company’s financial risk and its sources of funds. McWilliams and Siegel (2000) claimed that a high debt ratio implies high financial risk and will affect the company’s funding operations, the incentives toward performing CSR, and the company’s financial performance in the future. This research follows the model of Waddock and Graves (1997) by calculating the debt ratio as total liabilities divided by total assets and expects to see a negative correlation between the debt ratio and a company’s CSR performance.

Information Transparency (Disclosure Level)

The data used to evaluate information transparency is the evaluation results of the information disclosure and transparency ranking system created by the Securities and Futures Institute and organized by the TEJ. From 2007 to 2010, the company evaluation results were announced by five levels: A+ class, A class, B class, C class, and C− class. Since 2011, the levels have been adjusted to six: A++ class, A+ class, A class, B class, C class, and C− class. According to the description of the ranking process provided by the information disclosure and transparency ranking system, companies in the A+ class received grades above 80; after 2011, to subdivide the rankings in more detail, companies in the A++ class got grades above 85, and companies in the A+ class received grades above 80 but under 85. To ensure the data are unified, A++ grades and A+ grades from 2011 to 2014 will be considered as the same grade to make the data align with the rankings of 2007–2010. In this research, the range of the information transparency from 1 to 5, in which companies in the A++ class or A+ class receive the highest score of 5, ones in the A-class receive 4, ones in the B class receive 3, ones in the C class receive 2, and the ones in the C− class receives the lowest score of 1.

Loss

Ashbaugh-Skaife et al. (2006) believed that a company’s profitability strongly connects with its default risk. When a company suffers a loss, the default risk increases, this research follows the settings of the abovementioned study by creating the variable LOSS. LOSS is one if the company suffers a loss in the current period and the previous one and 0 otherwise.

General Manager Turnover

According to previous work, the number of general manager changes negatively correlates with company performance in the stock market (Weisbach, 1988). This result implies when the gap between the company’s operational performance and expected revenue increases that, the probability of changing the general manager increases and default risk thus increases. This research measures the general manager turnover variable by the number of general managers changes within the past 3 years.

Research Method

Before applying regression analysis to test the hypotheses, this research first uses regression to predict missing data in the CSR Award rankings and to transfer the censored data. The process goes as follows.

Predicting the Rankings Outside of the Top 40 in the CSR Award List

From 2007 to 2014, Common Wealth Magazine announced only the rankings of the top 40 companies for the CSR Award. The ones in the 41st place and lower were not open to the public, so the data of the top 40 companies were censored. Under this setting, a company with a ranking must be in the top 40, and a company without a ranking must be higher than 40. In order to predict the rankings after number 41, this research applies the Tobit regression model (Tobin, 1958). The Tobit regression model is designed to deal with censored data, in which all independent variables correspond to observations, while the dependent variables only have corresponding partial observations. When the proportion of observations is high, the expected value of the error term may not be 0 (Powell, 1986).

We first apply the Tobit model to get regression coefficients. This research sets the reciprocal of CSR rankings as the dependent variable. As discussed in previous works, the independent variables include factors that influence CSR rankings. The factors include debt ratio, institutional investors’ holding ratio, the size of directors and supervisors, the number of independent directors and supervisors, company size, ROA, and insiders’ shareholding. In this stage, the Tobit model goes as follows:

\begin{array}{l} EXP_{RANK}_{i t} = β_{0} + β_{1} DEBT_{RATIO}_{i t} + β_{2} INSTI_{RATIO}_{i t} + \\ β_{3} BOR_{SIZE}_{i t} + β_{4} INDEP_{RATIO}_{i t} + \\ β_{5} COM_{SIZE}_{i t} + β_{6} {ROA}_{i t} + \\ β_{7} INTER_{SHARE}_{i t} + ε_{i t} \end{array}

(1)

In the Tobit model above, the definitions of the variables are: EXP_RANK_it is the expected rank of company i in year t; DEBT_RATIO_it is the debt ratio of company i in year t; INSTI_RATIO_it is the institutional investors’ shareholding ratio of company i in year t; BRO_SIZE_it is the size of directors and supervisors of company i in year t; INDEP_RATIO_it is the proportion of independent directors and supervisors of company i in year t; COM_SIZE_it is the size of company i in year t; RAO_it is the ROA ratio of company i in year t; INTER_SHARE_it is insiders’ shareholding of company i in year t; and ε_it represents the error term.

Second, we build a prediction model with the regression coefficients and forecast the rankings outside the top 40 in the CSR award rankings. The results of the Tobit regression model appear in Table 1. The institutional investors’ shareholding ratio and ROA positively correlate with CSR rankings, and the results are statistically significant. Insiders’ shareholding negatively correlates with CSR rankings, and the results are statistically significant. The effects of other variables are not statistically significant. Thus, this research only includes institutional investors’ shareholding ratio, ROA ratio, and insiders’ shareholding in the CSR ranking prediction model. With the results in Table 1, the prediction model is

\begin{array}{l} EXP_\hat{{RANK}_{i t}} = 0.100867 + 0.003292 \times INSTI_{RATIO}_{i t} + \\ 0.004394 \times {ROA}_{i t} + (- 0.001503) \times \\ INTER_{SHARE}_{i t} \end{array}

(2)

Table 1.

Tobit Regression Model on the Reciprocal of CSR Rankings.

Variable	Coefficient	Std. Error	p Value
Constant	0.100867*	0.052084	.0528
Debt ratio	−0.000911	0.000633	.1503
Institutional investors shareholding	0.003292***	0.000936	.0004
Board size	−0.002491	0.003122	.4250
Independent directors	−0.006721	0.007371	.3619
Company size	2.37E-11	3.57E-11	.5074
ROA	0.004394***	0.001376	.0014
Insiders’ shareholding	−0.001503**	0.000603	.0126

Notes: *** is 0.01 significance, ** is 0.05 significance, and * is 0.1 significance.

Third, we generate a complete CSR award-ranking variable (RANK). As mentioned in the first step of this section, the dependent variable $EXP_\hat{{RANK}_{i t}}$ is the reciprocal of CSR rankings. In order to observe the predicted rankings, the reciprocal of $EXP_\hat{{RANK}_{i t}}$ is used to represent the rankings. Since all samples being predicted have actual rankings >40, this research adds 40 to all the predicted values, rounds them to the nearest whole number, and finally gets the predicted CSR Award rankings.

In the original dataset, the number of observations with rankings is 239, and the number of observations without rankings and required predictions is 288. Table 2 presents the descriptive statistics of the actual rankings and the predicted rankings.

Table 2.

Descriptive Statistics of the Original and Predicted CSR Rankings.

Statistics	Predicted Rankings	Original Rankings	Combined Rankings
Average	47.01	20.55	35.01
Median	46.00	21.00	44.00
Maximum	88.00	40.00	88.00
Minimum	41.00	1.00	1.00
SD	4.20	11.55	15.62
N	288	239	527

The result of Table 2 shows that the original rankings have a minimum ranking of 1 and a maximum ranking of 40. The average ranking is 21, the predicted rankings have a minimum ranking of 41, a maximum ranking of 88, and an average ranking of 47. The average ranking of the combined rankings is 35.

Empirical Model Settings

Based on the model structure mentioned in the previous section, this research builds three models, Model 3, Model 4, and Model 5, to test the hypotheses set in the ‘hypotheses’ construction’ section.

\begin{array}{l} {RANK}_{i t} = β_{0} + β_{1} INFOR_{LEVEL}_{i t} + β_{2} BOR_{SIZE}_{i t} + \\ β_{3} INDEP_{RATIO}_{i t} + β_{4} INSTI_{RATIO}_{i t} + \\ β_{5} INTER_{SHARE}_{i t} + β_{6} COM_{SIZE}_{i t} + \\ β_{7} {ROA}_{i t} + β_{8} DEBT_{RATIO}_{i t} + u_{1 i t} \end{array}

(3)

\begin{array}{l} {TCRI}_{i t} = α_{0} + α_{1} INFOR_{LEVEL}_{i t} + α_{2} BOR_{SIZE}_{i t} + \\ α_{3} INDEP_{RATIO}_{i t} + α_{4} INSTI_{RATIO}_{i t} + \\ α_{5} INTER_{SHARE}_{i t} + α_{6} COM_{SIZE}_{i t} + \\ α_{7} {ROA}_{i t} + α_{8} DEBT_{RATIO}_{i t} + \\ α_{9} {LOSS}_{i t} + α_{10} {MANA}_{i t} + u_{2 i t} \end{array}

(4)

\begin{array}{l} {TCRI}_{i t} = γ_{0} + γ_{1} {RANK}_{i t} + γ_{2} BOR_{SIZE}_{i t} + \\ γ_{3} INDEP_{RATIO}_{i t} + γ_{4} INSTI_{RATIO}_{i t} + \\ γ_{5} INTER_{SHARE}_{i t} + γ_{6} COM_{SIZE}_{i t} + \\ γ_{7} {ROA}_{i t} + γ_{8} DEBT_{RATIO}_{i t} + \\ γ_{9} {LOSS}_{i t} + γ_{10} {MANA}_{i t} + u_{3 i t} \end{array}

(5)

In the models above, the explanations of the variables are as follows: INFOR_LEVEL_it is the information transparency of company i in year t; RANK_it is the CSR ranking of company i in year t; TCRI_it is the credit rating of company i in year t, BOR_SIZE_it is the size of directors and supervisors of company i in year t; COM_SIZE_it is the size of company i in year t; INDEP_RATIO_it is the proportion of independent directors and supervisors of company i in year t; INSTI_RATIO_it is the institutional investors’ shareholding ratio of company i in year t; INTER_SHARE_it is insiders’ shareholding of company i in year t; DEBT_RATIO_it is the debt ratio of company i in year t; ROA_it is the ROA ratio of company i in year t; MANA_it is the number of general manager changes of company i in year t; LOSS_it equals to 1 if company i suffered a loss in year t or year t-1 and 0 otherwise; and u_1it, u_{2it and} u_1it represent the error term.

Correlation Analysis

According to Salahuddin and Islam (2008), when processing panel data, the Pearson correlation coefficient matrix must first be applied to test if a multi-collinearity issue exists. We later use the variance inflation factor (VIF) to test the results. This research uses the Pearson correlation coefficient matrix to test the correlation among the variables as the fundamental of regression analysis. As presented in Table 3, the correlation between any two variables is <0.7. Haan (2002) states that multi-collinearity arises when VIF is larger than 10, leading to the estimator losing efficiency (Maddala, 2005). The VIF testing results, appearing in Table 4, all have values <2, implying no multi-collinearity issue.

Table 3.

Correlation Coefficients.

	1	2	3	4	5	6	7	8	9	10	11	12
1. Information transparency	1
2. CSR ranking	−0.14***	1
3. Credit rating	−0.17***	0.26***	1
4. ROA	0.07	−0.24***	−0.51***	1
5. Manager change	−0.13**	0.06	0.05	−0.05	1
6. Loss	−0.14***	−0.12**	−0.41***	0.41***	−0.16***	1
7. Insiders’ shareholding	−0.01	0.02	0.04	−0.08	0.10**	−0.12**	1
8. Institutional investors shareholding	−0.06	−0.25***	−0.31***	0.30***	−0.08*	0.14***	0.19***	1
9. Independent directors	−0.17***	−0.06	−0.19***	0.26***	0.06	0.11**	−0.15***	0.26***	1
10. Debt ratio	0.20***	0.08*	0.21***	−0.41***	0.05	−0.21***	−0.07	−0.26***	−0.11**	1
11. Company size	0.10*	−0.14***	−0.32***	0.10***	0.07	0.0223	−0.12**	0.12**	0.23***	0.15***	1
12. Board size	0.11**	−0.02	−0.12**	−0.00	−0.11	0.09*	0.23***	0.17***	−0.02	0.03	0.01	1

Note: *** is 0.01 significance, ** is 0.05 significance, and * is 0.1 significance.

Table 4.

VIFs for Variables in the Three Models.

	Model 3	Model 4	Model 5
Information transparency	1.1410	1.2138	–
CSR award rankings	–	–	1.1319
Board size	1.0950	1.1300	1.1365
Independent directors	1.2490	1.2556	1.1648
Institutional investors shareholding	1.2878	1.3065	1.2871
Insiders’ shareholding	1.1597	1.2112	1.2170
Company size	1.1419	1.1506	1.1552
ROA	1.4054	1.6397	1.5510
Debt ratio	1.3906	1.3885	1.3191
Loss	–	1.3192	1.2860
Manager change	–	1.1017	1.0571

Weak-Instrument Test and Testing for Endogeneity

After generating the complete data, this research then performs the Durbin-Wu-Hausman test (Durbin, 1954; Hausman, 1978; Wu, 1973) to examine for endogeneity. This research first runs the weak-instrument test and endogeneity test since endogeneity may exist in variables such as information transparency or CSR ranking, leading to the estimation being biased. Referring to the method of previous research (Ebbes et al., 2016; Li et al., 2015), this present study uses information transparency and the CSR ranking of the previous year as instrument variables: the instrument variable of INFOR_LEVEL_it is INFOR_LEVEL_{it – 1}, or the information transparency of years t−7 to t−1; the instrument variable of RANK_it is RANK_{it – 1}; and the CSR ranking is for years t−7 to t−1. This research then applies two-stage least squares (2SLS) method, in which the model’s endogenous variables are expressed as functions of the exogenous variables. The information transparency and the CSR ranking are expressed as

\begin{array}{l} INFOR_{LEVEL}_{i t} = β_{0} + β_{1} INFOR_{LEVEL}_{i t - 1} + β_{2} BOR_{SIZE}_{i t} + \\ β_{3} INDEP_{RATIO}_{i t} + β_{4} INSTI_{RATIO}_{i t} + \\ β_{5} INTER_{SHARE}_{i t} + β_{6} COM_{SIZE}_{i t} + \\ β_{7} {ROA}_{i t} + β_{8} DEBT_{RATIO}_{i t} + ε_{1 i t} \end{array}

(6)

\begin{array}{l} INFOR_{LEVEL}_{i t} = α_{0} + α_{1} INFOR_{LEVEL}_{i t - 1} + α_{2} BOR_{SIZE}_{i t} + \\ α_{3} INDEP_{RATIO}_{i t} + α_{4} INSTI_{RATIO}_{i t} + \\ α_{5} INTER_{SHARE}_{i t} + α_{6} COM_{SIZE}_{i t} + \\ α_{7} {ROA}_{i t} + α_{8} DEBT_{RATIO}_{i t} + \\ α_{9} {LOSS}_{i t} + α_{10} {MANA}_{i t} + ε_{2 i t} \end{array}

(7)

\begin{array}{l} {RANK}_{i t} = γ_{0} + γ_{1} {RANK}_{i t - 1} + γ_{2} BOR_{SIZE}_{i t} + \\ γ_{3} INDEP_{RATIO}_{i t} + γ_{4} INSTI_{RATIO}_{i t} + \\ γ_{5} INTER_{SHARE}_{i t} + γ_{6} COM_{SIZE}_{i t} + \\ γ_{7} {ROA}_{i t} + γ_{8} DEBT_{RATIO}_{i t} + γ_{9} {LOSS}_{i t} + \\ γ_{10} {MANA}_{i t} + ε_{3 i t} \end{array}

(8)

All the variables follow explanations in previous sections, where ε_1it, ε_2it and ε_3it, represent the residual terms. To further check if the instrumental variable estimation can solve the endogenous problem in Models 3, 4, and 5, this research performs weak-instrument and endogeneity tests.

Two assumptions—instrument relevance and instrument exogeneity—must be satisfied when performing instrumental variable estimation (Stock et al., 2002). If the assumption of instrument relevance is not satisfied, then the instrument variables have little correlation with the endogenous variables, potentially leading to the issue of weak instruments. The weak instrument test checks β_1, α_{1 and} γ₁ in Models 6, 7, and 8. If this research can reject the null hypothesis, then the models are free from weak instrument issues; instrument exogeneity refers to the covariances between the instrument variables (INFOR_LEVEL_{it – 1 and} RANK_{it – 1}), and the error terms (u_1it, u_{2it and} u_3it) are 0 (Wooldridge, 2002).

According to Stock and Yogo (2005), when the Cragg-Donald F-stat is larger than 10.3, the test rejects the null hypothesis of the instrument variable being a weak instrument. Table 5 shows the test results, where the Cragg-Donald F-stat values of INFOR_LEVEL_{it – 1 and} RANK_{it – 1} in Model 7 and Model 8 are, respectively, 171.3084 and 197,370.9, implying that the instrument variables are not weak instruments. The coefficients of the instrument variables are both positive and the results are statistically significant, implying that the instrument variables of this research satisfy the assumption of instrument relevance.

Table 5.

Stage-1 Estimation Result of 2SLS and Weak Instrument Test.

Variable	Model 6	Model 7	Model 8
Variable	INFOR_LEVEL_it	INFOR_LEVEL_it	RANK_it
Constant	1.03***	1.50***	0.31*
INFOR_LEVEL_it–1	0.62***	0.59***
RANK_it–1			1.00***
BOR_SIZE_it	0.02**	0.03**	0.005
INDEP_RATIO_it	−0.13***	−0.13***	−0.03
INSTI_RATIO_it	−0.004	−0.005	−0.004
INTER_SHARE_it	−0.004	−0.004	0.0008
COM_SIZE_it	0.0000	0.0000	−0.000
ROAit	0.02***	0.03***	0.006
DEBT_RATIO_it	0.004	0.004*	−0.0006
LOSS_it		−0.46***	−0.16
MANA_it		−0.12**	0.03
Adjusted R-squared	0.40	0.42	1.00
F-statistic	34.02***	29.11***	22784.57***
Cragg-Donald F-stat	0.0395	171.3084	197370.9

Note: *** is 0.01 significance, ** is 0.05 significance, and * is 0.1 significance.

This research then uses the Durbin-Wu-Hausman test to examine the instrument exogeneity assumption. The first step is to express the endogenous variables as functions of the exogenous variables and find the residual of the estimators, INFOR_LEVEL_residhat and Rank_residhat. The second step is to substitute these residuals into Models 3, 4, and 5 and treat the residuals as independent variables. The adjusted models are expressed as follows:

\begin{array}{l} {RANK}_{i t} = β_{0} + β_{1} INFOR_{LEVEL}_{i t} + β_{2} BOR_{SIZE}_{i t} + \\ β_{3} INDEP_{RATIO}_{i t} + β_{4} INSTI_{RATIO}_{i t} + \\ β_{5} INTER_{SHARE}_{i t} + β_{6} COM_{SIZE}_{i t} + \\ β_{7} {ROA}_{i t} + β_{8} DEBT_{RATIO}_{i t} + \\ η_{1} INFOR_LEVEL_residhat + ξ_{1 i t} \end{array}

(9)

\begin{array}{l} {TCRI}_{i t} = α_{0} + α_{1} INFOR_{LEVEL}_{i t} + α_{2} BOR_{SIZE}_{i t} + \\ α_{3} INDEP_{RATIO}_{i t} + α_{4} INSTI_{RATIO}_{i t} + \\ α_{5} INTER_{SHARE}_{i t} + α_{6} COM_{SIZE}_{i t} + \\ α_{7} {ROA}_{i t} + α_{8} DEBT_{RATIO}_{i t} + α_{9} {LOSS}_{i t} + \\ α_{10} {MANA}_{i t} + η_{2} INFOR_LEVEL_residhat + ξ_{2 i t} \end{array}

(10)

\begin{array}{l} {TCRI}_{it} = γ_{0} + γ_{1} {RANK}_{i t} + γ_{2} BOR_{SIZE}_{i t} + \\ γ_{3} INDEP_{RATIO}_{i t} + γ_{4} INSTI_{RATIO}_{i t} + \\ γ_{5} INTER_{SHARE}_{i t} + γ_{6} COM_{SIZE}_{i t} + \\ γ_{7} {ROA}_{i t} + γ_{8} DEBT_{RATIO}_{i t} + γ_{9} {LOSS}_{i t} + \\ γ_{10} {MANA}_{i t} + η_{3} Rank_residhat + ξ_{3 i t} \end{array}

(11)

All the variables follow explanations in previous sections, where ξ_1it, ξ_2it and ξ_3it are the residuals. This research then uses t-statistics to test the null hypotheses H_{0: η1 = 0,} H_{0: η2 = 0 and} H_{0: η3 = 0}. If the t-statistics reject the null hypotheses, then INFOR_LEVEL_residhat and Rank_residhat are endogenous variables. The endogenous test results appear in Table 6. The t-statistics of these variables are both not statistically significant, implying that the information transparency and the CSR ranking are not significantly endogenous variables.

Table 6.

Stage-2 Estimation Result of 2SLS, Overidentification Test, and Endogenous Test.

Variable	Model 9	Model 10	Model 11
Variable	RANK_it	TCRI_it	TCRI_it
Constant	58.49***	5.1177***	2.43***
INFOR_LEVEL_it	−5.17	0.2372
RANK_it			0.0014
BOR_SIZE_it	0.1583	−0.0476**	0.0276
INDEP_RATIO_it	−0.1526	0.0653	0.0525*
INSTI_RATIO_it	−0.4125***	−0.0145**	−0.0009
INTER_SHARE_it	0.0673	0.0026	0.0299***
COM_SIZE_it	−0.0000	−0.0000***	−0.0000
ROA_it	−0.2348	−0.0737***	−0.0292***
DEBT_RATIO_it	0.0441	0.0025	0.0177***
LOSS_it		−0.8491**	−0.2139*
MANA_it		0.0122	0.0442
R-squared	0.1377	0.4389	0.9160
Over identification test χ² = N*R²	52.6014***	166.782***	418.612***
Endogenous test t-statistics	0.6032	−1.0654	−0.2696

Notes: 1. *** is 0.01 significance, ** is 0.05 significance, and * is 0.1 significance. 2. Over identification test χ² (d.f = 1) = 2.7062 at 0.1 significance, χ² (d.f = 1) = 3.841 at 0.05 significance, and χ² (d.f = 1) = 6.635 at 0.01 significance.

Empirical Results

This research finds no endogenous problem and applies regression models to find the relationships among CSR performance, credit rating, and information transparency. The dataset used in this research is multi-period panel data. In order to avoid omitted variable bias, panel data analysis is applied in the model other than ordinary least squares (OLS) regression. Thus, this research runs statistical tests on the three models—OLS, fixed-effects model, and random-effects model—to select the best model for this research. Table 7 presents the testing results.

Table 7.

Statistical Test Results of Regression Model Selection.

H0	OLS		OLS		One-way Random
H1	One-way Fixed		One-way Random		One-way Fixed
Model	F-Test	Model Fit	LM-Test	Model Fit	Hausman-Test	Model Fit
Model 3	3.26***	Fixed effect	68.52***	Random effect	19.02**	Fixed effect
Model 4	37.31**	Fixed effect	767.83***	OLS	58.42***	Fixed effect
Model 5	36.35***	Fixed effect	776.70***	OLS	37.79***	Fixed effect

Note: *** is 0.01 significance and ** is 0.05 significance.

At the 1% significance level, the fixed-effects model is a better fit since the constants of the three models are different, based on the F-test statistic and p value (Baltagi, 2008). At the 1% significance level, the random-effects model is a better fit for Model 3 since the constant of Model 3 is a random variable; OLS is a better fit for Model 4 and Model 5 since the constants of the models are not random variables, based on the LM-test statistic and p value (Breusch & Pagan, 1980). The F-test and LM-test results show that the fixed-effects model and the random-effect model are better fits than the OLS model. Thus, the next step is to select between the fixed effect model and the random effect model. At the 1% significance level, the results of the Hausman test (Hausman, 1978) and the p value shown in the random effect model that there is a correlation between the constants of the three models and the independent variables. Therefore, the fixed-effects model is a better fit for this research.

The next step is applying a statistical test to check the fixed effect on cross-sectional and longitudinal data. The results appear in Table 8. In Models 3, 4, and 5, fixed effects can be observed on cross-sectional data, and the F-test results on longitudinal data of the three models all do not reach the 10% level of significance and cannot reject the null hypothesis of no fixed effect on longitudinal data. Based on the results, this research will only analyze fixed effects on cross-sectional data via the regression model.

Table 8.

Fixed Effect Testing on Cross-Sectional Data and Longitudinal Data.

Model	Effect Testing	F-Test	p Value	Fixed Effect
Model 3	Cross-sectional	3.51	<.0001	Yes
Model 3	Longitudinal	0.47	.86	No
Model 4	Cross-sectional	41.85	<.0001	Yes
Model 4	Longitudinal	0.64	.72	No
Model 5	Cross-sectional	39.49	<.0001	Yes
Model 5	Longitudinal	1.00	.42	No

The dependent variable of Model 3 is the CSR ranking. Table 9 shows that information transparency has a significantly positive effect on CSR performance, represented by a lower number of CSR ranking. When a company has a higher information transparency, the company’s CSR ranking will also be higher. This result is in line with Hypothesis 1; that is, when a company discloses information, leading to an increase in information transparency, it will improve the company’s CSR performance.

The regression result about Model 4 in Table 9 shows that information transparency has a negative effect on credit rating, and the result is statistically significant, showing that the result supports Hypothesis 2, which states that when a company manager discloses more information to the public, the company’s default risk will decrease, leading to a better credit rating and lower credit risk. The result also implies that when a company discloses more information, shareholders will expect that the possibility of the company hiding negative information is lower, causing the default risk to decrease and credit rating to get better.

The regression result about Model 5 in Table 9 shows that CSR ranking has a negative effect on credit rating, and the result is statistically significant, meaning that the result validates Hypothesis 3. The possible reasons behind this correlation are that when a company is active in CSR activities, it can avoid potential sanctions from stakeholders both legally and economically and can more easily obtain external resources, leading to lower credit risk and a better credit rating.

Table 9.

Regression Models and Empirical Results.

Variable	Model 3	Model 4	Model 5
Dependent Variable	CSR Ranking	Credit Rating	Credit Rating
Constant	49.3427***	7.0085***	5.5824***
Independent variables
INFOR_LEVEL	−1.9281***	−0.2618***
RANK			0.0103***
Control variables
BOR_SIZE	0.0954	−0.0238	−0.0580***
INDEP_RATIO	0.4553	−0.0064	0.0224
INSTI_RATIO	−0.3419***	−0.0199***	−0.0200***
INTER_SHARE	0.0339	0.0001	0.0137***
COM_SIZE	−0.0000**	−0.0000***	−0.0000***
ROA	−0.3347***	−0.0514***	−0.0567***
Debt ratio	−0.0007	0.0059	0.0009
LOSS		−1.3720***	−0.9328***
MANA		−0.0695	−0.0837

Note: *** is 0.01 significance, ** is 0.05 significance, and * is 0.1 significance.

Concluding Remarks

Conclusion and Suggestions

If a company can increase its information transparency level, be more open about financial and non-financial information to the public, and more actively communicate with stakeholders, then the issue of information asymmetry can be reduced. The increased information transparency level can lower the risk of moral hazard or managers who practice self-interested or illegal actions, thus decreasing a company’s default risk, which improves the company’s credit performance. A company can also disclose information through its CSR performance in order to create a positive company image and increase corporate value. This action can decrease unsystematic risk and improve the company’s credit rating.

The abovementioned findings align with the agency theory, overinvestment view, and the risk mitigation viewpoints implying that if a company discloses more information to the public, then agency costs can fall, and its CSR performance can improve, leading to a better credit rating and implying a lower default risk.

The empirical results of this research show that when a company manager discloses information to shareholders or stakeholders and does his job via corporate governance, he is also taking on the responsibilities of practicing CSR by reducing the agency cost generated by the supervision system due to information asymmetry. Since CSR performance and information transparency are important factors in evaluating credit rating institutions, strong performance in these two metrics leads to the company’s credit rating improving and further makes investors more confident in the company. A company evaluated as low-risk and receiving investments from shareholders can boost its disclosure of financial and non-financial information. The three factors of CSR, information transparency, and credit ratings go hand in hand and have strong relationships.

Based on the empirical analysis results, this research suggests that companies build proper information disclosure channels, provide related information to stakeholders in real-time, and work to improve their image through CSR. These activities can help decrease a firm’s credit rating and improve its social image. A better corporate image leads to a higher level of information transparency and thus creates a positive cycle, continually reaching the sustainable development goals.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

References

Aggarwal

R. K.

, & Nanda

(2004). Access, common agency, and board size. SSRN Journal. 10.2139/ssrn.571801

Aguilera-Caracuel

, & Guerrero-Villegas

(2018). How corporate social responsibility helps MNEs to improve their reputation. The moderating effects of geographical diversification and operating in developing regions. Corporate Social Responsibility and Environmental Management, 25(4), 355–372

Ahn

, Bonsall

S. B.

, & Van Buskirk

(2019). Do managers withhold bad news from credit rating agencies? Review of Accounting Studies, 24, 972–1021.

Al-Shammari

M. A.

, Banerjee

S. N.

, & Rasheed

A. A.

(2021). Corporate social responsibility and firm performance: A theory of dual responsibility. Management Decision, 60(6), 1513–1540. 10.1108/MD-12-2020-1584

Aldamen

, Duncan

, Kelly

, & McNamara

(2020). Corporate governance and family firm performance during the global financial crisis. Accounting & Finance, 60(2), 1673–1701.

Ang

J. S.

, & Brau

J. C.

(2002). Firm transparency and the costs of going public. Journal of Financial Research, 25(1), 1–17.

Anwar

, & Malik

J. A.

(2020). When does corporate social responsibility disclosure affect investment efficiency? A new answer to an old question. Sage Open, 10(2), 2158244020931121.

Arab

, Gholamrezapoor

, & Toraj

(2020). The mediating effect of information asymmetry and agency costs on the relationship between CSR and investment efficiency. Advances in Mathematical Finance and Applications, 5(2), 149–166.

Ashbaugh-Skaife

, Collins

D. W.

, & LaFond

(2006). The effects of corporate governance on firms’ credit ratings. Journal of Accounting and Economics, 42(1/2), 203–243.

10.

Attig

, El Ghoul

, Guedhami

, & Suh

(2013). Corporate social responsibility and credit ratings. Journal of Business Ethics, 117, 679–694.

11.

Bae

S. C.

, Chang

, & Yi

H. C.

(2018). Corporate social responsibility, credit rating, and private debt contracting: New evidence from syndicated loan market. Review of Quantitative Finance and Accounting, 50, 261–299.

12.

Baginski

S. P.

, Hassell

J. M.

, & Kimbrough

M. D.

(2002). The effect of legal environment on voluntary disclosure: Evidence from management earnings forecasts issued in US and Canadian markets. The Accounting Review, 77(1), 25–50.

13.

Baltagi

B. H.

(2008). Forecasting with panel data. Journal of Forecasting, 27(2), 153–173.

14.

Bannier

C. E.

, Hirsch

C. W.

, & Wiemann

(2012). Do credit ratings affect firm investments? The monitoring role of rating agencies. SSRN. https://dx-doi-org-s.web.bisu.edu.cn/10.2139/ssrn.2135483

15.

Bao

, Kim

, Mian

G. M.

, & Su

(2019). Do managers disclose or withhold bad news? Evidence from short interest. The Accounting Review, 94(3), 1–26.

16.

Barnea

, & Rubin

(2010). Corporate social responsibility as a conflict between shareholders. Journal of Business Ethics, 97(1), 71–86.

17.

Bénabou

, & Tirole

(2010). Individual and corporate social responsibility. Economica, 77(305), 1–19.

18.

Benlemlih

, Ge

, & Zhao

(2021). Undervaluation and non-financial information: Evidence from voluntary disclosure of CSR news. Journal of Business Finance & Accounting, 48(5/6), 785–814.

19.

Berle

A. A.

, & Means

G. C.

(1932). The modern corporation and private property. Transaction Publishers.

20.

Bonsall IV

S. B.

(2014). The impact of issuer-pay on corporate bond rating properties: Evidence from Moody’s and S&P’s initial adoptions. Journal of Accounting and Economics, 57(2/3), 89–109.

21.

Boutin-Dufresne

, & Savaria

(2004). Corporate social responsibility and financial risk. Journal of Investing, 13(1), 57–66.

22.

Breusch

T. S.

, & Pagan

A. R.

(1980). The Lagrange multiplier test and its applications to model specification in econometrics. The Review of Economic Studies, 47(1), 239–253.

23.

Brickley

J. A.

, Lease

R. C.

, & Smith Jr

C. W.

(1988). Ownership structure and voting on antitakeover amendments. Journal of Financial Economics, 20, 267–291.

24.

Brown

, Beekes

, & Verhoeven

(2011). Corporate governance, accounting and finance: A review. Accounting & Finance, 51(1), 96–172.

25.

Bushman

R. M.

, Smith

A. J.

, & Wittenberg-Moerman

(2010). Price discovery and dissemination of private information by loan syndicate participants. Journal of Accounting Research, 48(5), 921–972.

26.

Chau

G. K.

, & Gray

S. J.

(2002). Ownership structure and corporate voluntary disclosure in Hong Kong and Singapore. The International Journal of Accounting, 37(2), 247–265.

27.

Chen

S. S.

, Chen

H. Y.

, Chang

C. C.

, & Yang

S. L.

(2013). How do sovereign credit rating changes affect private investment? Journal of Banking & Finance, 37(12), 4820–4833.

28.

Cherian

, Safdar

S. M.

, Tran

D. K.

, Hwang

, Khanh

T. H. T.

, & Ahmed

(2020). The strength of CEOs’ influence on CSR in Chinese listed companies. New insights from an agency theory perspective. Sustainability, 12(6), 2190.

29.

Chi

, Liu

, & Wang

(2009). What affects accounting conservatism: A corporate governance perspective. Journal of Contemporary Accounting & Economics, 5(1), 47–59.

30.

Choy

, Gray

, & Ragunathan

(2006). Effect of credit rating changes on Australian stock returns. Accounting & Finance, 46(5), 755–769.

31.

Diamond

D. W.

, & Verrecchia

R. E.

(1991). Disclosure, liquidity, and the cost of capital. Journal of Finance, 46(4), 1325–1359.

32.

Djankov

, La Porta

, Lopez-de-Silanes

, & Shleifer

(2008). The law and economics of self-dealing. Journal of Financial Economics, 88(3), 430–465.

33.

Duffie

, & Lando

(2001). Term structures of credit spreads with incomplete accounting information. Econometrica, 69(3), 633–664.

34.

Durbin

(1954). Errors in variables. Review of the International Statistical Institute, 22, 23–32.

35.

Ebbes

, Papies

, & van Heerde

H. J.

(2016). Dealing with endogeneity: A nontechnical guide for marketing researchers. In Handbook of market research (pp. 1–37). Springer.

36.

Fabozzi

F. J.

, Ng

P. W.

, & Tunaru

D. E.

(2022). The impact of corporate social responsibility on corporate financial performance and credit ratings in Japan. In Risks related to environmental, social and governmental, issues (ESG) (pp. 3–19). Springer Nature.

37.

Fama

E. F.

, & Jensen

M. C.

(1983). Separation of ownership and control. Journal of Law and Economics, 26(2), 301–325.

38.

Ferrell

, Liang

, & Renneboog

(2016). Socially responsible firms. Journal of Financial Economics, 122(3), 585–606.

39.

Fitch Ratings. (2004). Credit policy special report, evaluating corporate governance: The bondholders’ perspective. Fitch Ratings.

40.

Fombrun

, & Shanley

(1990). What’s in a name? Reputation building and corporate strategy. Academy of Management Journal, 33(2), 233–258.

41.

Fuente

J. A.

, García-Sanchez

I. M.

, & Lozano

M. B.

(2017). The role of the board of directors in the adoption of GRI guidelines for the disclosure of CSR information. Journal of Cleaner Production, 141, 737–750.

42.

Global Reporting Initiative. (2020). Sustainability disclosure database—Data legend. Global Reporting Initiative.

43.

Godfrey

P. C.

(2005). The relationship between corporate philanthropy and shareholder wealth: A risk management perspective. Academy of Management Review, 30(4), 777–798.

44.

Goss

, & Roberts

G. S.

(2011). The impact of corporate social responsibility on the cost of bank loans. Journal of Banking and Finance, 35(7), 1794–1810.

45.

Gurun

U. G.

, & Butler

A. W.

(2012). Don’t believe the hype: Local media slant, local advertising, and firm value. The Journal of Finance, 67(2), 561–598.

46.

Haan

C. T.

(2002). Statistical methods in hydrology (2nd ed., pp. 128–160). Blackwell.

47.

Hassen

R. B.

, & Militaire

M. A. A.

(2020). The impact of corporate social responsibility on executive compensation. Open Access Library Journal, 7(8), 1.

48.

Hart

O. D.

, & Moore

(1995). Debt and seniority: An analysis of the role of hard claims in constraining management. American Economic Review, 85(3), 567–585.

49.

Hauge

(2019). Should the African lion learn from the Asian tigers? A comparative-historical study of FDI-oriented industrial policy in Ethiopia, South Korea and Taiwan. Third World Quarterly, 40(11), 2071–2091.

50.

Hausman

J. A.

(1978). Specification tests in econometrics. Econometrica, 46(6), 1251–1271.

51.

Healy

P. M.

, Hutton

A. P.

, & Palepu

K. G.

(1999). Stock performance and intermediation changes surrounding sustained increases in disclosure. Contemporary Accounting Research, 16(3), 485–520.

52.

Healy

P. M.

, & Palepu

K. G.

(2001). Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical disclosure literature. Journal of Accounting and Economics, 31(1–3), 405–440.

53.

Hermalin

B. E.

, & Weisbach

M. S.

(2003). Boards of directors as an endogenously determined institution: A survey of the economic literature. Economic Policy Review, 9, 7–26.

54.

Hillman

A. J.

, & Dalziel

(2003). Boards of directors and firm performance: Integrating agency and resource dependence perspectives. Academy of Management Review, 28(3), 383–396.

55.

Hong

, & Kacperczyk

(2010). Competition and bias. Quarterly Journal of Economics, 125(4), 1683–1725.

56.

Hsiao

(1986). Analysis of panel data (1st ed.). Cambridge University Press.

57.

Huang

, Sun

, & Chu

(2022). Can we-media information disclosure drive listed companies’ innovation?—From the perspective of financing constraints. China Finance Review International, 12(3), 477–495.

58.

Hutton

A. P.

, Miller

G. S.

, & Skinner

D. J.

(2003). The role of supplementary statements with management earnings forecasts. Journal of Accounting Research, 41(5), 867–890.

59.

Jamali

, Safieddine

A. M.

, & Rabbath

(2008). Corporate governance and corporate social responsibility synergies and interrelationships. Corporate Governance: An International Review, 16(5), 443–459.

60.

Jensen

M. C.

, & Meckling

W. H.

(1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305–360.

61.

Jensen

M. C.

, & Ruback

R. S.

(1983). The market for corporate control: The scientific evidence. Journal of Financial Economics, 11(1–4), 5–50.

62.

Jiraporn

, Jiraporn

, Boeprasert

, & Chang

(2014). Does corporate social responsibility (CSR) improve credit ratings? Evidence from geographic identification. Financial Management, 43(3), 505–531.

63.

, & Harjoto

M. A.

(2011). Corporate governance and firm value: The impact of corporate social responsibility. Journal of Business Ethics, 103(3), 351–383.

64.

, & Na

(2012). Does CSR reduce firm risk? Evidence from controversial industry sectors. Journal of Business Ethics, 110, 441–456.

65.

Johnson

R. A.

, & Greening

D. W.

(1999). The effects of corporate governance and institutional ownership types on corporate social performance. Academy of Management Journal, 42(5), 564–576.

66.

Jorion

, Liu

, & Shi

(2005). Informational effects of regulation FD: Evidence from rating agencies. Journal of Financial Economics, 76(2), 309–330.

67.

Keim

G. D.

(1978). Corporate social responsibility: An assessment of the enlightened self-interest model. Academy of Management Review, 3(1), 32–39.

68.

Kothari

S. P.

, Shu

, & Wysocki

P. D.

(2009). Do managers withhold bad news? Journal of Accounting Research, 47(1), 241–276.

69.

Kotler

, & Lee

(2005). Best of breed: When it comes to gaining a market edge while supporting a social cause, “corporate social marketing” leads the pack. Social Marketing Quarterly, 11(3/4), 91–103.

70.

Kumar

, Langberg

, & Sivaramakrishnan

(2012). Voluntary disclosures, corporate control, and investment. Journal of Accounting Research, 50(4), 1041–1076.

71.

Lau

L. J.

(2020). The competitive advantage of Taiwan. In The competitive advantages of far eastern business (pp. 90–111). Routledge.

72.

Lee

C. H.

, Huang

S. Y.

, Barnes

F. B.

, & Kao

(2010). Business performance and customer relationship management: The effect of IT, organisational contingency and business process on Taiwanese manufacturers. Total Quality Management, 21(1), 43–65.

73.

, Lu

, Mittoo

U. R.

, & Zhang

(2015). Board independence, ownership concentration and corporate performance - Chinese evidence. International Review of Financial Analysis, 41, 162–175.

74.

Lin

, Hung

P. H.

, Chou

D. W.

, & Lai

C. W.

(2019). Financial performance and corporate social responsibility: Empirical evidence from Taiwan. Asia Pacific Management Review, 24(1), 61–71.

75.

Maddala

G. S.

(2005). Introduction to econometrics (3rd ed.). John Wiley & Sons Ltd.

76.

Martínez‐Ferrero

, Garcia‐Sanchez

I. M.

, & Cuadrado‐Ballesteros

(2015). Effect of financial reporting quality on sustainability information disclosure. Corporate Social Responsibility and Environmental Management, 22(1), 45–64.

77.

McConnell

J. J.

, & Servaes

(1995). Equity ownership and the two faces of debt. Journal of Financial Economics, 39(1), 131–157.

78.

McWilliams

, & Siegel

(2000). Corporate social responsibility and financial performance: Correlation or misspecification? Strategic Management Journal, 21(5), 603–609.

79.

Mercer

(2005). The fleeting effects of disclosure forthcomingness on management’s reporting credibility. The Accounting Review, 80(2), 723–744.

80.

Micu

, Remolona

, & Wooldridge

(2004). The price impact of rating announcements: Evidence from the credit default swap market. BIS Quarterly Review, 2, 55–65.

81.

Milgrom

P. R.

(1981). Good news and bad news: Representation theorems and applications. Bell Journal of Economics, 12(2), 380–391.

82.

Nwude

E. C.

, & Nwude

C. A.

(2021). Board structure and corporate social responsibility: Evidence from developing economy. Sage Open, 11(1), 2158244020988543.

83.

Park

B. J.

, & Lee

K. H.

(2021). The sensitivity of corporate social performance to corporate financial performance: A “time-based” agency theory perspective. Australian Journal of Management, 46(2), 224–247.

84.

Pawlina

, & Renneboog

(2005). Is investment‐cash flow sensitivity caused by agency costs or asymmetric information? Evidence from the UK. European Financial Management, 11(4), 483–513.

85.

Phang

S. Y.

, & Hoang

(2021). Does positive CSR increase willingness to invest in a company based on performance? The incremental role of combined assurance. Accounting & Finance, 61(4), 5631–5654.

86.

Pittman

(2008). Moody’s, S&P defer cuts on AAA subprime, hiding loss. Bloomberg News Service March, 11.

87.

Powell

J. L.

(1986). Symmetrically trimmed least squares estimation for Tobit models. Econometrica, 54(6), 1435–1460.

88.

Porter

M. E.

, & Kramer

M. R.

(2006). Strategy and society: The link between corporate social responsibility and competitive advantage. Harvard Business Review, 84(12), 78–92.

89.

Preston

L. E.

, & O’bannon

D. P.

(1997). The corporate social-financial performance relationship: A typology and analysis. Business & Society, 36(4), 419–429.

90.

Salahuddin

, & Islam

M. R.

(2008). Factors affecting investment in developing countries: A panel data study. Journal of Developing Areas, 42(1), 21–37.

91.

Sengupta

(1998). Corporate disclosure quality and the cost of debt. The Accounting Review, 73(4), 459–474.

92.

Sethuraman

(2019). The effect of reputation shocks to rating agencies on corporate disclosures. The Accounting Review, 94(1), 299–326.

93.

Shleifer

, & Vishny

R. W.

(1986). Large shareholders and corporate control. Journal of Political Economy, 94(3), 461–488.

94.

Skinner

D. J.

(1994). Why firms voluntarily disclose bad news. Journal of Accounting Research, 32(1), 38–60.

95.

Starks

L. T.

(2009). EFA keynote speech: Corporate governance and corporate social responsibility: What do investors care about? What should investors care about? Financial Review, 44, 461–468.

96.

Stock

J. H.

, Wright

J. H.

, & Yogo

(2002). A survey of weak instruments and weak identification in generalized method of moments. Journal of Business & Economic Statistics, 20(4), 518–529.

97.

Stock

, & Yogo

(2005). Testing for weak instruments in linear IV regression. In Andrews

D. W. K.

(Ed.), Identification and inference for econometric models (pp. 80–108). Cambridge University Press.

98.

Tahir

, Nazir

M. S.

, Qamar

M. A. J.

, & Boyer

M. M.

(2022). Ineffective implementation of corporate governance? A call for greater transparency to reduce agency cost. Managerial and Decision Economics, 43(5), 1528–1547.

99.

Tobin

(1958). Estimation of relationships for limited dependent variables. Econometrica, 26(1), 24–36.

100.

Waddock

S. A.

, & Graves

S. B.

(1997). The corporate social performance–financial performance link. Strategic Management Journal, 18(4), 303–319.

101.

Weisbach

M. S.

(1988). Outside directors and CEO turnover. Journal of Financial Economics, 20, 431–460.

102.

Wooldridge

J. M.

(2002). Econometric analysis of cross section and panel data. MIT Press.

103.

D. M.

(1973). Alternative tests of independence between stochastic regressors and disturbances. Econometrica, 41(4), 733–750.

104.

Yermack

(1996). Higher market valuation of companies with a small board of directors. Journal of Financial Economics, 40(2), 185–211.

105.

E. P. Y.

, Guo

C. Q.

, & Luu

B. V.

(2018). Environmental, social and governance transparency and firm value. Business Strategy and the Environment, 27(7), 987–1004.

106.

, & Zhang

(2011). Do lenders value corporate social responsibility? Evidence from China. Journal of Business Ethics, 104, 197–206.

107.

(2005). Accounting transparency and the term structure of credit spreads. Journal of Financial Economics, 75(1), 53–84.